INDEX

Explanations

get it, have it, much of

New Auto-Interp

Configuration

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

他们

-1.22

 ones

-1.22

 they

-1.08

 mereka

-1.04

NUMBER

-1.01

 number

-1.00

它们

-0.99

 them

-0.98

 которых

-0.98

They

-0.97

POSITIVE LOGITS

 stuff

1.77

 some

1.71

 much

1.50

 Much

1.47

Much

1.46

 none

1.27

it

1.24

 that

1.20

 None

1.16

 Some

1.14

Activations Density 0.099%